Friday Facts #442 - Flip, Flow, and Fresh Paint

[Eulenberg]

Trains are unfortunately (almost) useless and quality won't save them.
Bigger acceleration and top speed is nice, but they are still inferior to belts.
Belts effectively teleport items, with a small downside of buffer capacity.
But they do not require all the setup with the stations and their management, schedules, intrusive tracks in the middle of the base, refuelling, congestion, and so on.
It is too much work for no reward.

It sounds like you're trying to build trains around your factory instead of building a factory around the trains.
Also you can rename stations and if you give two stations the same name trains can go to either one of them (can't go to both at the same time because trains can only be in one place at a time), this makes trains capable of n-to-n logistics as opposed to a belt's 1-to-1 logistical capabilities.
And you can name your train schedules so when building a new iron plate train you can just pick the schedule and boom, instant iron train, edits in one train also apply to all others with the same schedule.
Since trains deliver large bursts of supplies instead of a constant stream like belts you also wanna buffer the resources, so instead of directly train to belt you go train to buffer to belt or in reverse for loading trains, you'll also wanna buffer trains in waiting bays, so they can arrive before they're needed.
For refueling set up some refueling stations and use interrupts to have trains go refuel whenever they're low on fuel or use bots to transport fuel to every station
Congestion only really happens if you mess up with signals or if you try to have more than 1 train on a bi-directional track.
Tracks and trains are general purpose infrastructure that will be used to transport many different resources and will remain useful even after the original reasons they were built are gone. Compare it to belts where once either end of the line is gone the belt loses it's purpose.

I think what he means, and i agree with him the buffs are not enough to bring back a train meta in the endgame, it not worth to build around trains like in bigger mod packs, one reason is the liquid buss the other is the fact that in 1.x a train would carry multiple blue belts worth of items 2-3 depending on stack size if not even more per wagon if im correct, but in space age, even the 100 slot wagon will struggle to carry 1 single fully stacked green belt. also belts are way more optimised than they used to be, especially with gapless/back pressured belt, also the base game has little to no use case for true n to n, or even lazy n to n, so that the trains are just big overhead.

the 100 slot wagon are in the game for quite some time already, the quality scaling on them is just disabled, i already tested them in a playtough back when they put them in, and while it makes it more rewarding if you REALLY like trains, they still vastly inferior to green stacked belts.

[emty]

It sounds like you're trying to build trains around your factory instead of building a factory around the trains.
Also you can rename stations and if you give two stations the same name trains can go to either one of them (can't go to both at the same time because trains can only be in one place at a time), this makes trains capable of n-to-n logistics as opposed to a belt's 1-to-1 logistical capabilities.
And you can name your train schedules so when building a new iron plate train you can just pick the schedule and boom, instant iron train, edits in one train also apply to all others with the same schedule.
Since trains deliver large bursts of supplies instead of a constant stream like belts you also wanna buffer the resources, so instead of directly train to belt you go train to buffer to belt or in reverse for loading trains, you'll also wanna buffer trains in waiting bays, so they can arrive before they're needed.
For refueling set up some refueling stations and use interrupts to have trains go refuel whenever they're low on fuel or use bots to transport fuel to every station
Congestion only really happens if you mess up with signals or if you try to have more than 1 train on a bi-directional track.
Tracks and trains are general purpose infrastructure that will be used to transport many different resources and will remain useful even after the original reasons they were built are gone. Compare it to belts where once either end of the line is gone the belt loses it's purpose.

Yes, I know that stuff, I used trains a lot in 1.0.
The problem is, again, that Factorio, does not require all of this n-to-n stuff.
The production blocks are fixed and just require a steady supply of resources.
It is easy to achieve in a self-contained 1-to-1 way even without balancing between belts (and fluids are even easier).
That's just how the current "physics" of the world work.
This is sad, because while trains are interesting, they don't give enough benefits to justify their complexity.



Also, unrelated to above, I reread FFF and I am confused now:

Asteroid reprocessing recipe returning asteroid chunks.
Research for increasing the reprocessing yield.
Being able to put Quality modules in the reprocessing recipes.
<...>
So in 2.1 placing Quality modules in the asteroid reprocessing recipes is disallowed.

It's unclear what is meant by *reprocessing* recipes, because they mention research affecting reprocessing *yield*, but the *yield* only affect crushing recipes.
So I don't get it.

If quality is allowed on crushing recipes, then it's not the biggest deal, since 30 lvl research still gives an 1 -> 0.8 asteroid reduction rate on basic crushing with the downside of generating a lot of trash byproducts like ice.
If quality is disallowed on crushing as well, that's totally not fun and just means more constraints for plastic and stone.

But normal reprocessing recipes with 1 -> 0.8 reduction without ANY productivity research are indeed too strong.